Abstract
Normal colour vision relies on the absorption of light by the short-, medium-, and long-wavelength-sensitive cones. If one or more cone types are absent, or their function is compromised (e.g., diabetic retinopathy, macular degeneration, Leber Congenital Amaurosis), colour vision is affected. Most of the available colour vision tests (with some exceptions, e.g. Simunovic et al., 1998; Arden & Wolf, 2004; Barbur, 2004) can only assess the degree of colour vision in observers with visual acuity above 0.1 (6/60). To measure the residual colour vision retained by observers with low vision, we have developed a new test that consists of circles of random luminance presented on a computer screen. A small sub-set of circles delineate a large 5-degree square that varies in saturation. Using a 2AFC paradigm, observers indicated whether the coloured square appeared on the left- or the right-hand side of the screen. During the test, the chromaticity of the square changed along several directions of the CIE (1976) Luv space. Using a staircase procedure, we measured the minimum saturation required for each observer to discriminate the square from the achromatic background (white point). We found that for adult observers with normal vision, chromatic discrimination thresholds were equally distant from the white point in all directions. More importantly, affected adults and children, who were unable to perform various standardised tests (including the Cambridge Colour Test) due to their low visual acuity, could successfully complete our test. However, their thresholds had overall larger departures from the white point and were better characterised by ellipses with axes of asymmetrical length, with the main axis being parallel to the protan, the deutan or the tritan confusion axes. The test is a promising tool for measuring and monitoring changes in colour vision due to the progression of a disease or its improvement after treatment.
Meeting abstract presented at VSS 2013